The Floor That Wasn't

The bot this week

It's real now — well, real on paper. The book went live 25 days ago (on paper, $10,000 start) and has been rebalancing daily straight into one of the ugliest weeks crypto has had in months.

Equity	$10,606.36
Return since launch	+6.06%
vs. holding BTC	−16.64%
Max drawdown	−4.53%
Funding (annualized)	+2.26%
Net positioning	net short (−$2,443.34)

Here's the part that matters: the book is +6.06% while simply holding BTC would be −16.64%. It went net short into the selloff and got paid for it. That's not a forecast — it's the whole thesis showing up in real time: make money, or at least don't bleed, when the market falls.

Same window, holding BTC would have returned −16.64%. The book is beating buy-and-hold by roughly +22.71% — almost entirely by being net short into a falling tape, not by calling the top.

How it's positioned

The book runs short perps against long spot on each asset — a defensive, low-net-exposure stance the trend filter flipped into as momentum rolled over. Drawdown so far is a trivial −4.53%, and the funding leg is even paying a small carry (+2.26% annualized).

The honest caveat

This is 25 days of paper, not a track record. A green first week proves nothing — the backtest already told us the edge is risk reduction, not outperformance: out-of-sample, fixed parameters, it returned about the same as buy-and-hold (Sharpe 0.93 vs 0.62) with roughly a third of the drawdown. A calm week and a violent one are both just single data points. Real capital goes in around July; until then, every number here is paper and public.

The $58K Print: Leverage, Macro, and a Very Thin Bid

Start with the number that defined the week: $58,000. Bitcoin's drop there wasn't dramatic in isolation — it's roughly where the market traded in late 2024. But the context made it brutal. This is a 21-month low, hit during what was supposed to be a post-ETF, institutional-adoption bull cycle. The costume came off.

The mechanism was layered. First, US PCE inflation data came in at a three-year high, immediately killing any residual hope of a Fed pivot. Rate-sensitive risk assets sold off globally — South Korea's Kospi dropped over 8% and triggered a circuit breaker. Crypto didn't decouple; it amplified. The PCE print landed on a market that was already fragile, and within an hour, roughly $600 million in crypto positions were liquidated.

Then came the ETF data, which made the institutional narrative look shaky in real time. US spot Bitcoin ETFs recorded $696.3 million in outflows in a single session — June's worst daily number, and the largest since May. Year-to-date, spot BTC ETF flows are now net negative by $4.6 billion. These are the same vehicles that were supposed to anchor institutional demand. Right now they're functioning as an exit door.

Strategy's preferred shares added a sideshow with teeth. STRC slid 26% below par — a significant move for an instrument that's supposed to track near $100 — while MSTR shares hit a 16-month low. For a company that has made levered Bitcoin accumulation its entire identity, watching the preferred stack crack is not a footnote. It's a signal that the market is repricing the risk embedded in that structure.

Options traders at Anchorage Digital noted that the market is positioned defensively but not catastrophically — near-term hedges are heavy, though no one's pricing a full Strategy blowup. Bear flag technicals, however, pointed to a potential target of $54,000 or lower if the breakdown confirmed. Bitcoin's power-law model frames $58K as a normal cycle low. Futures market positioning says the bottom may not be in. Both can be true: the selloff is structurally explainable, and it may not be done.

ETH's Quiet Capitulation

Ether had a worse week than Bitcoin on a relative basis — and the symbolism was hard to miss. ETH fell to $1,558, revisiting support levels last seen in October 2023 and April 2025. The move was sharp enough for Tether's market cap to flip Ether's — a milestone that tells you something real about where confidence is sitting. Four Ethereum OG wallets that had held 37,602 ETH since 2018 finally sold this week, locking in an estimated $27 million in profit after watching $150 million in unrealized gains evaporate at the peak. Eight years of conviction. Sold near the bottom of the range.

Against that backdrop, Sharplink made a contrarian call: the firm bought 5,000 ETH — its first purchase in eight months — days after helping fund Ethlabs, a nonprofit spun out of former Ethereum Foundation researchers. It's a small position but a directional signal from an entity that has been publicly building an ETH treasury thesis. Meanwhile, USDe's circulating supply has shrunk 70% from its October peak of $14 billion, underscoring how much the broader Ethena ecosystem shrank with the bull market.

Framework's $400M Bet: Crypto Isn't Enough Anymore

Framework Ventures closed a $400 million fourth fund — and the mandate tells you where smart crypto money thinks the next decade lives. The fund will invest across crypto, AI, robotics, and energy. This isn't a pivot away from crypto; Framework says it will continue those investments. But the expansion is deliberate: the firm also promoted Rajiv Patel-O'Connor to general partner as it builds out the non-crypto vertical. It's the same logic Galaxy Digital has been running — crypto infrastructure firms are increasingly being valued for AI adjacency, not just digital asset exposure. Galaxy's stock surge this year is exhibit A. Framework is placing the same bet with fresh capital.

Polymarket, AscendEX, and the Week's Trust Deficit

Two separate platform integrity stories ran in parallel this week, and together they were a reminder that the infrastructure layer still has real failure modes. Polymarket suffered a $2.9 million theft after attackers injected a malicious script into its frontend via a compromised third-party dependency. The platform said it contained the breach and will refund affected users. The timing is painful: Polymarket has been positioning itself as a mainstream onboarding layer — roughly 60% of World Cup bettors on the platform were first-time crypto users. That's a real audience. A frontend hack in that window is a reputational problem, not just a security incident.

AscendEX is a different and potentially more serious story. Multiple users reported delayed withdrawals after on-chain analyst ZachXBT flagged potential liquidity issues and publicly called on the exchange to disclose its reserves. No resolution was confirmed by week's end. Separately, Europol froze €41 million in crypto tied to a global infostealer malware takedown, disrupting operations including SocGholish and StealC — tools designed specifically to harvest wallet credentials.

Regulation: MiCA Bites, Singapore Flags Hyperliquid, Senate Stalls

Three regulatory threads worth tracking. Binance will restrict EU onboarding and services from July 1 after failing to secure MiCA authorization from any EU member state — withdrawals remain available, but new user flows into the EU are effectively throttled. Singapore's MAS added Hyperliquid to its Investor Alert List, a reminder that operating without a license in the region carries consequences even for leading decentralized exchanges. Multicoin's price target of $319 for HYPE by 2028 and its "everything exchange" thesis may be right, but it doesn't exempt the protocol from national regulatory frameworks. In the US, the Senate's crypto legislative push is facing a narrowing window — housing bill turmoil is threatening the timeline for a broader crypto bill to advance in July.

What to Watch

• $58K–$54K range: Bear flag technicals point to $54,000 as the next meaningful target if Bitcoin fails to reclaim $60K on a daily close.
• ETF flows: A seventh consecutive day of outflows would mark the longest streak of 2026 — watch whether institutional demand returns or the bleed extends into July.
• MiCA July 1 deadline: Binance's EU service restrictions go live. Watch for user migration data and whether other exchanges use the opening to gain market share.
• Senate crypto bill: July is the stated target window. Any slippage past August materially reduces the odds of passage before the legislative calendar compresses.